The Center of Innovation in Datatech and Artificial Intelligence (CIDAI) and the Cybersecurity Agency of Catalonia (ACC) are pleased to invite you to the presentation of the results of an innovative project that addresses the improvement in access and consultation of information within large unstructured documentary volumes. This project, based on the implementation of artificial intelligence models, has as its objective objective to respond effectively and contextually to the queries made by the Agency’s staff, through the real-time information retrieval.
Through the recovery augmented generation (RAG) methodology, the power of pre-trained language models is combined with an advanced information retrieval system, thus allowing to improve the accuracy of the answers and provide explicit references that support the decisions made. This solution offers an answer to the difficulty of managing PDF documents and other dense text files in unstructured formats, and will allow for a more agile and precise query in large data sets.
CIDAI’s impact projects promotes the Government through the Secretariat of Digital Policies.
This project, led by Eurecat, has worked with a modular system and highly parameterized which has facilitated efficient modifications to the various layers of the system. From content indexing, to retrieval and generation of answers, searching improve the overall efficiency of the information management system in the Agency.
In this session, CIDAI, in collaboration with the Cybersecurity Agency of Catalonia, will present the results and knowledge acquired during its execution.
